A Convex Surrogate Operator for General Non-Modular Loss Functions

Abstract : Empirical risk minimization frequently employs convex surrogates to underlying discrete loss functions in order to achieve computational tractability during optimization. However, classical convex surrogates can only tightly bound modular loss functions, sub-modular functions or supermodular functions separately while maintaining polynomial time computation. In this work, a novel generic convex surrogate for general non-modular loss functions is introduced, which provides for the first time a tractable solution for loss functions that are neither super-modular nor submodular. This convex surro-gate is based on a submodular-supermodular decomposition for which the existence and uniqueness is proven in this paper. It takes the sum of two convex surrogates that separately bound the supermodular component and the submodular component using slack-rescaling and the Lovász hinge, respectively. It is further proven that this surrogate is convex , piecewise linear, an extension of the loss function, and for which subgradient computation is polynomial time. Empirical results are reported on a non-submodular loss based on the Sørensen-Dice difference function, and a real-world face track dataset with tens of thousands of frames, demonstrating the improved performance, efficiency, and scalabil-ity of the novel convex surrogate.
Type de document :
Communication dans un congrès
The 19th International Conference on Artificial Intelligence and Statistics, May 2016, Cadiz, Spain. Proceedings of the 19th International Conference on Artificial Intelligence and Statistics
Liste complète des métadonnées

https://hal.inria.fr/hal-01299519
Contributeur : Jiaqian Yu <>
Soumis le : mardi 12 avril 2016 - 14:01:57
Dernière modification le : vendredi 12 janvier 2018 - 10:55:24
Document(s) archivé(s) le : mercredi 13 juillet 2016 - 10:29:33

Fichiers

aistats2016.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01299519, version 1
  • ARXIV : 1604.03373

Citation

Jiaqian Yu, Matthew Blaschko. A Convex Surrogate Operator for General Non-Modular Loss Functions. The 19th International Conference on Artificial Intelligence and Statistics, May 2016, Cadiz, Spain. Proceedings of the 19th International Conference on Artificial Intelligence and Statistics. 〈hal-01299519〉

Partager

Métriques

Consultations de la notice

218

Téléchargements de fichiers

111